Canadians are comfortable letting AI handle routine banking, but not major, long-term financial decisions.
TD’s annual AI Insights Report found that just over half of Canadians are comfortable with financial institutions using AI to track spending (55%), calculate credit scores (53%), and half say they’d use AI themselves for budgeting.
For quick, transactional moments like checking balances, resetting passwords, or getting answers about fees, many Canadians actually prefer AI over waiting for a human.
There’s a limit on this trust, though.
Only one in three Canadians trust AI over their parents for financial advice. Seventy-one per cent still place greater confidence in human intelligence overall, and for major decisions with long-term impact like retirement planning, product approvals, and financial planning, most want a person in the room.
Reasons for the distrust are specific, with 52% citing concerns about errors in high-stakes situations. The top three situations are inaccurate information (61%), privacy and security risks (55%), and lack of accountability (54%).
It’s a similar story with organizations. Employees at all levels dip their toes in AI, but when it gets a number wrong in an important document or project, the whole thing is shelved.
Consumer sentiment and enterprise deployment are living the same AI reality.
What would move the needle, according to respondents, is data protection (56%), accountability for errors (55%), and human oversight (52%). Fifty-nine per cent say they’re comfortable with banks using AI as long as meaningful human oversight exists.
“AI has the potential to enhance how we serve clients by improving speed, convenience and personalization, but it is still human connection that builds trust and provides the understanding, reassurance and advice clients need during important financial moments,” says Jayme Martin, district vice president, Greater Hamilton South at TD Bank Group.
What this shows tech leaders is that customers will use these tools, and they might even be OK with others using them on their behalf (eg. customer service or organizing their data), but they want someone accountable behind it.
They want to know a human is behind it, and that someone will answer if and when it goes wrong.
Final Shots
- Canadians are comfortable with AI for routine banking tasks, but 71% still trust human judgment overall
- The top trust barriers are inaccuracy, privacy risk, and lack of accountability, not the technology itself
- Human oversight is the single biggest trust-builder, with 59% say they’re comfortable with AI in banking when a human is meaningfully in the loop